July 2012 - kernelnewbies@kernelnewbies.org - Kernelnewbies Mailing Lists

Using GDB for debugging netlink communication
by Ranjan Sinha July 22, 2012

July 22, 2012

Hi, I have a multi-threaded application that communicates with a kernel module using netlink sockets. One of the threads in user mode application works as a server and kernel module works as a client. Roughly the kernel code is as follows: -------------------------8<------------------------------------------------8<------------------------------------- timeout = 3500; netlink_unicast(); __set_current_state(TASK_INTERRUPTIBLE); wait: timeout = schedule_timeout(timeout); __set_current_state(TASK_RUNNING); if (!timeout) { printk(KERN_ERR "No response received\n"); return -1; } if (message_status != UPDATED) { printk(KERN_ERR "Somebody woke us up before we got a reply. Time left %d\n", timeout); __set_current_state(TASK_INTERRUPTIBLE); goto wait; } -------------------------8<------------------------------------------------8<------------------------------------- The message_status variable is updated in the netlink callback when the user mode application replies to this message. So basically the idea is to send a message and then wait at max timeout jiffies for the reply. Now, using gdb, if I add a break point in any function that is called by netlink server thread in user mode, the break point is never hit and the kernel log is flooded with messages like -------------------------8<------------------------------------------------8<------------------------------------- Somebody woke us up before we got a reply. Time left 3499 Somebody woke us up before we got a reply. Time left 3499 Somebody woke us up before we got a reply. Time left 3499 Somebody woke us up before we got a reply. Time left 3499 .. .. Somebody woke us up before we got a reply. Time left 3498 -------------------------8<------------------------------------------------8<------------------------------------- Until I finally get -------------------------8<------------------------------------------------8<------------------------------------- No response received -------------------------8<------------------------------------------------8<------------------------------------- What is causing the kernel thread to wake up from the timeout and how should I debug the user mode code? -- Best Regards, Ranjan PS: I am using 2.6.32-71.el6.x86_64 on RHEL 6.0

2 1

Re: ipv6 address comparison
by Arif Hossain July 21, 2012

July 21, 2012

On Sat, 2012-07-21 at 15:36 +0200, Sven-Haegar Koch wrote: > On Sat, 21 Jul 2012, Arif Hossain wrote: > > > My module uses ipv6_addr_cmp() from <linux/ipv6.h>. > > > > <linux/ipv6.h> > > static inline int ipv6_addr_cmp(const struct in6_addr *a1, const struct > > in6_addr *a2) > > { > > return memcmp(a1, a2, sizeof(struct in6_addr)); > > } > > > > I've included <linux/ipv6.h>. But when i try to build the module i get > > following error: > > > > net/netfilter/xt_ipaddr.c:81: error: implicit declaration of function > > ‘ipv6_addr_cmp’ > > > > Although before the call to ipv6_addr_cmp, i have a casting to struct > > ipv6hdr *iph. So the definitions from ipv6.h should be available to my > > module. > > In my 3.2.0 source tree here the definition is in include/net/ipv6.h, > not in include/linux/ipv6.h > > c'ya > sven-haegar > > -- > Three may keep a secret, if two of them are dead. > - Ben F. Thanks for the reply. Now its working. cheers

1 0

ipv6 address comparison
by Arif Hossain July 21, 2012

July 21, 2012

Hi, My module uses ipv6_addr_cmp() from <linux/ipv6.h>. <linux/ipv6.h> static inline int ipv6_addr_cmp(const struct in6_addr *a1, const struct in6_addr *a2) { return memcmp(a1, a2, sizeof(struct in6_addr)); } I've included <linux/ipv6.h>. But when i try to build the module i get following error: net/netfilter/xt_ipaddr.c:81: error: implicit declaration of function ‘ipv6_addr_cmp’ Although before the call to ipv6_addr_cmp, i have a casting to struct ipv6hdr *iph. So the definitions from ipv6.h should be available to my module. cheers. aft

2 1

Device Mapper Query
by Satyajeet July 20, 2012

July 20, 2012

Does the dmsetup <dm-dev> <table-file> command, calls the constructor* * dm_ctr_fn?? If yes what are the argv passed in the dm_ctr_fn, are they the args you pass to dmsetup?? Is this constructor a callback from dm layer? when it is invoked and how it is invoked? Replies will be appreciated. Regards, satyajeet

3 2

UDP packets are getting dropped because of invalied checksum though checksum offload is enabled in kernel-2.6.32
by Murali Annamneni July 20, 2012

July 20, 2012

Hi All, I have written a kernel module using netfilter framework. In my code I'll do some modifications to the header & payload parts of UDP packet in pre & post hooks. After modifying the udp packet, I'm computing ip & udp checksum. This computation is consuming some cpu cycles. Now, I want to offload this computation to the Ethernet device. For this I have commented the checksum computation in my code and enabled the checksum offload parameters of my Ethernet device. While testing, I found that kernel is dropping the packets because of invalid checksums for ip & udp packets. I have few doubts regarding checksum offload 1) Shall I need to do anything in my kernel code. 2) I'm using the below command to enable checksum offload ethtool -K eth0 rx on tx on Is there anything I need to do apart from this ? I'm testing my kernel code in kernel-2.6.32. Thanks & Regards Murali Annamneni ::DISCLAIMER:: ---------------------------------------------------------------------------------------------------------------------------------------------------- The contents of this e-mail and any attachment(s) are confidential and intended for the named recipient(s) only. E-mail transmission is not guaranteed to be secure or error-free as information could be intercepted, corrupted, lost, destroyed, arrive late or incomplete, or may contain viruses in transmission. The e mail and its contents (with or without referred errors) shall therefore not attach any liability on the originator or HCL or its affiliates. Views or opinions, if any, presented in this email are solely those of the author and may not necessarily reflect the views or opinions of HCL or its affiliates. Any form of reproduction, dissemination, copying, disclosure, modification, distribution and / or publication of this message without the prior written consent of authorized representative of HCL is strictly prohibited. If you have received this email in error please delete it and notify the sender immediately. Before opening any email and/or attachments, please check them for viruses and other defects. ----------------------------------------------------------------------------------------------------------------------------------------------------

3 3

Problem with module spanning from multiple files
by aleksey July 18, 2012

July 18, 2012

Hello. I want to build kernel module which consist from two files. test.c: #include <linux/kernel.h> #include <linux/module.h> int my_module_init(void) { pr_emerg("Hello, world - this is the kernel speaking\n" ) ; return 0; } MODULE_DESCRIPTION("test driver" ) ; MODULE_LICENSE("GPL v2" ) ; MODULE_VERSION("0.1" ) ; module_init(my_module_init) ; module_exit(my_module_exit) ; test_sub.c: #include <linux/kernel.h> #include <linux/module.h> void my_module_exit() { pr_emerg("Short is the life of a kernel module\n" ) ; } Makefile: obj-m = test.o test-objs = test_sub.o all: $(MAKE) -C $(KDIR) M=$(shell pwd) modules The module is building without any warnings. When i execute insmod none of message is appear. It's look like my_module_init function does not executed at all.

3 5

setup_thread_stack
by Francesco Scali July 18, 2012

July 18, 2012

Hi, I'm totally new to kernel code, just trying to understand the basics of the core subsystems. During the do_fork call, there is a call to dup_task_struct, which in turn calls setup_thread_stack. What's this call supposed to do? Inside I can see a call to the "task_thread_info" macro, but I don't quite understand it: #define task_thread_info(task) ((struct thread_info *)(task)->stack) can a task_struct be cast to a thread_info? perhaps my C knowledge should be improved..I know :( Sorry for the trivial question..just trying to understand :) Regards, Francesco

3 2

good explanation of __read_mostly, __init, __exit macros,
by Aft nix July 17, 2012

July 17, 2012

Hi, The macro expansion of __read_mostly : #define __read_mostly __attribute__((__section__(".data..read_mostly")) This one is from cache.h __init: #define __init __section(.init.text) __cold notrace from init.h __exit: #define __exit __section(.exit.text) __exitused __cold notrace After searching through net i have not found any good explanation of what is happening there. Additonal question : I have heard about various "linker magic" employed in kernel development. Any information regarding this will be wonderful. cheers -- -aft

7 10

Makefiles in Linux
by Satyajeet July 17, 2012

July 17, 2012

How to set Module Version and module Release date in the makefile? Is it by the way of CFLAGS? Regards, satya

2 1

login blocked - [/bin/sh: can't access tty; job control turned off]
by pcuser p July 17, 2012

July 17, 2012

Hi, Login prompt at the end of kernel bootup is blocked & getting the messasge "/bin/sh: can't access tty; job control turned off" I've tried with the options in busybox FAQ,nothing worked out. Details ls -l /dev/console crw-rw-rw- 1 root root 204, 64 Dec 31 17:00 /dev/console ls -l /dev/ttySAC0 crw-rw-rw- 1 root root 204, 64 Dec 31 17:00 /dev/ttySAC0 /etc/inittab: ::sysinit:/etc/init.d/rcS ttySAC0::respawn:/sbin/getty -L ttySAC0 115200 vt100 # GENERIC_SERIAL In ::sysinit:/etc/init.d/rcS i'm doing : exec /bin/sh setsid sh -c 'exec sh </dev/ttySAC0 >/dev/ttySAC0 2>&1' by the completion of /etc/inittabscript,I'm ending in shell prompt with the messsge "/bin/sh: can't access tty; job control turned off" But I can run all commads from the shell prompt (Except login,if I give system hangs) I feel the information, I provided may not be very clear,but this is my first time working in linux. plz suggest a way to get login working.

1 0